Back to Home

My Journey

The roles, impact, and milestones that shaped how I build.

Atrium · Pfizer · United States
AI Engineer

Automating clinical trial documentation with AI

Jun 2025 – Aug 2025

Worked with Pfizer's AI team to automate how statisticians write analysis plans. Co-authored a peer-reviewed publication in Clinical Trials (SAGE). Built a RAG pipeline and an LLM-as-a-Judge system for hallucination detection.

  • Co-authored a peer-reviewed publication in Clinical Trials (SAGE) on AI-driven SAP authoring
  • Automated Statistical Analysis Plan generation, reducing drafting time by 60%
  • Evaluated chunking strategies, embedding models, rerankers, and HyDE — improved retrieval relevance by 18%
  • Built LLM-as-a-Judge evaluation with 82% precision and 78% recall
  • Conducted ablations improving generation consistency by 25% and reducing revision cycles by 33%
60%
Faster SAP
drafting
82%
Hallucination
detection precision
25%
Generation
consistency gain

Manager's recommendation

"He brought in new techniques for prompt engineering and automated evaluation that we did not have yet."
— Rogier Landman, Pfizer R&D

Tezo · 3 Years · India
Machine Learning Engineer

RAG chatbot over 1,000+ internal documents

Jul 2022 – Jul 2024
  • Built a RAG-powered chatbot over project charters, design specs, and status reports
  • Engineered semantic search with embeddings + vector DB, reducing lookup time by 61%
  • LLM-based summarization cut manual review time for managers by 37%
  • Containerized deployment with Docker + CI/CD for secure on-prem access
1k+
Internal docs
indexed
61%
Faster document
lookups
37%
Less manual
review time

Promoted in 1 year

Junior to Machine Learning Engineer, recognized for initiative and delivery.

Junior Machine Learning Engineer

Fraud detection & data infrastructure

Jul 2021 – Jul 2022
  • Migrated policy and claims data to Snowflake, reducing query times by 40%
  • Trained ML models for fraud detection, improving recall by 12%
  • Built BI dashboards adopted by 50+ users across actuarial and operations teams
12%
Fraud recall
improvement
Education

MS Applied Machine Learning

University of Maryland, College Park · 3.9 GPA
Coursework in deep learning, NLP, and inference optimization.
Aug 2024 – Present

3.9
GPA